Anthropic set to launch new model before IPO to counter GPT-6 Astra

According to Reuters, three sources indicated that Anthropic is exploring the launch of a new AI model to counter OpenAI’s rapid rollout of GPT-6 Astra. This development occurs as Anthropic prepares for its initial public offering, shortly after its CEO urged the broader AI industry to decelerate technological iteration.
Sources stated that introducing a new model aims to address competitive pressures. Previously, Anthropic CEO Dario Amodei called on the global AI sector to slow the release of new capabilities and properly manage safety risks. “We must slow down the speed of AI model capability iteration,” he wrote in a 3,800-word article published on September 12. The piece painted a concerning picture: numerous AI agents are crawling the internet, with development outpacing human control—a perspective shared by Sam Altman and Elon Musk.
Astra Gains Enterprise Market Share Over Claude
OpenAI’s GPT-6 Astra, launched on September 3, highlights computer operations, software engineering, cybersecurity, and professional office tools, quickly attracting business clients. Data from enterprise expense management platform Ramp shows Astra now represents roughly 13% of enterprise AI spending, compared to Anthropic’s Claude Fable at around 8%. On OpenRouter, users spent more on OpenAI models than Anthropic models last week, marking the first time OpenAI has led this metric in two and a half years. Consequently, some investors have reevaluated Anthropic’s standing as a top enterprise AI provider.
Nevertheless, Anthropic retains a revenue advantage: as of late July, its annualized revenue run rate surpassed $65 billion (approximately RMB 436.696 billion), up from about $9 billion at the end of 2025; the company forecasts revenue between $190 billion and $200 billion by 2028. OpenAI’s annualized revenue run rate in July exceeded $40 billion.
Key Challenges: Open-Source Models and Meta’s Strategic Shift
Investors believe a more significant threat stems from the growth of open-source, open-weight models, which can lower token usage costs and enable companies to build proprietary AI infrastructure. One source noted that Meta, formerly one of Anthropic’s key clients, plans to reduce its reliance on Anthropic’s models and instead develop more AI capabilities in-house.
A source revealed that while Anthropic discusses releasing a new version, it continues to assess the safety level of this next-generation model; the company is also weighing how to balance new model development with enhancing profitability. Regarding these reports, Anthropic declined to comment.
On the IPO timeline, Altman confirmed OpenAI will not go public in 2026; according to two sources, Anthropic may postpone its IPO until after the US November mid-term elections, whereas Reuters previously reported it might begin its IPO roadshow as early as mid-October.
Related article
Kimi K3.1 Set for Next-Month Launch With Three Reasoning Tiers and Million-Token Capacity
Moonshot is preparing to release the Kimi K3.1 model next month, featuring three adjustable reasoning tiers: Low, High, and Max. Leaked internal API data from X user @MaxForAI confirms these details, revealing model identifiers like "k3d1-agent" alon
How to fix Core Web Vitals for better SEO rankings
Top AI Tools for Converting Video and Audio to TextTable of Contents:IntroductionWeet.io - Convert Video to TextYouTube Subtitles and TranscriptionsFigo.com - Fastest Video to Text ConversionBlack Box - Screenshots of Video to Text ConversionYouTube
Xiaopeng Humanoid Robot Factory in Guangzhou Starts Trial Production, Eyes 2026 for Mass Output
Xiaopeng’s humanoid robot has entered small-scale trial production at its Guangzhou facility. The mass production line is currently undergoing final integration, signaling the start of the countdown to full-scale manufacturing.Previously, He Xiaopeng
Related Special Topic Recommendations
Comments (0)
0/500

According to Reuters, three sources indicated that Anthropic is exploring the launch of a new AI model to counter OpenAI’s rapid rollout of GPT-6 Astra. This development occurs as Anthropic prepares for its initial public offering, shortly after its CEO urged the broader AI industry to decelerate technological iteration.
Sources stated that introducing a new model aims to address competitive pressures. Previously, Anthropic CEO Dario Amodei called on the global AI sector to slow the release of new capabilities and properly manage safety risks. “We must slow down the speed of AI model capability iteration,” he wrote in a 3,800-word article published on September 12. The piece painted a concerning picture: numerous AI agents are crawling the internet, with development outpacing human control—a perspective shared by Sam Altman and Elon Musk.
Astra Gains Enterprise Market Share Over Claude
OpenAI’s GPT-6 Astra, launched on September 3, highlights computer operations, software engineering, cybersecurity, and professional office tools, quickly attracting business clients. Data from enterprise expense management platform Ramp shows Astra now represents roughly 13% of enterprise AI spending, compared to Anthropic’s Claude Fable at around 8%. On OpenRouter, users spent more on OpenAI models than Anthropic models last week, marking the first time OpenAI has led this metric in two and a half years. Consequently, some investors have reevaluated Anthropic’s standing as a top enterprise AI provider.
Nevertheless, Anthropic retains a revenue advantage: as of late July, its annualized revenue run rate surpassed $65 billion (approximately RMB 436.696 billion), up from about $9 billion at the end of 2025; the company forecasts revenue between $190 billion and $200 billion by 2028. OpenAI’s annualized revenue run rate in July exceeded $40 billion.
Key Challenges: Open-Source Models and Meta’s Strategic Shift
Investors believe a more significant threat stems from the growth of open-source, open-weight models, which can lower token usage costs and enable companies to build proprietary AI infrastructure. One source noted that Meta, formerly one of Anthropic’s key clients, plans to reduce its reliance on Anthropic’s models and instead develop more AI capabilities in-house.
A source revealed that while Anthropic discusses releasing a new version, it continues to assess the safety level of this next-generation model; the company is also weighing how to balance new model development with enhancing profitability. Regarding these reports, Anthropic declined to comment.
On the IPO timeline, Altman confirmed OpenAI will not go public in 2026; according to two sources, Anthropic may postpone its IPO until after the US November mid-term elections, whereas Reuters previously reported it might begin its IPO roadshow as early as mid-October.
Kimi K3.1 Set for Next-Month Launch With Three Reasoning Tiers and Million-Token Capacity
Moonshot is preparing to release the Kimi K3.1 model next month, featuring three adjustable reasoning tiers: Low, High, and Max. Leaked internal API data from X user @MaxForAI confirms these details, revealing model identifiers like "k3d1-agent" alon
How to fix Core Web Vitals for better SEO rankings
Top AI Tools for Converting Video and Audio to TextTable of Contents:IntroductionWeet.io - Convert Video to TextYouTube Subtitles and TranscriptionsFigo.com - Fastest Video to Text ConversionBlack Box - Screenshots of Video to Text ConversionYouTube
Xiaopeng Humanoid Robot Factory in Guangzhou Starts Trial Production, Eyes 2026 for Mass Output
Xiaopeng’s humanoid robot has entered small-scale trial production at its Guangzhou facility. The mass production line is currently undergoing final integration, signaling the start of the countdown to full-scale manufacturing.Previously, He Xiaopeng





Home






